Re: msvcp140: implement _To_byte

[Nikolay Sivov]

On 08.07.2017 22:33, Mikhail Paulyshka wrote:
> Signed-off-by: Mikhail Paulyshka
>
> dlls/msvcp140/msvcp140.spec | 2 +-
>  dlls/msvcp140/tests/msvcp140.c | 49 ++++++++++++++++++++++++++++++++++++++++++
>  dlls/msvcp90/locale.c | 10 +++++++++
>  3 files changed, 60 insertions(+), 1 deletion(-)

Hi, and thanks for the patch.

Some comments in addition to Stefan's reply:
> - static const WCHAR test_2[] = {0xfe95, 0xfea0, 0xfea6, 0}; //some arabic glyphs

These are characters, not glyphs. Also please use C-style comments - /**/.

msvcp140: implement _To_byte

[Henri Verbeet]

On 9 July 2017 at 12:58, Nikolay Sivov  wrote:
> On 08.07.2017 22:33, Mikhail Paulyshka wrote:
> > - static const WCHAR test_2[] = {0xfe95, 0xfea0, 0xfea6, 0}; //some arabic glyphs
>
> These are characters,

Well, kind of. These are actually from the Presentation Forms-B block.
I'm not sure why those would be especially relevant for the function
in question though.

msvcp140: implement _To_byte

[Stefan Dösinger]

Am 2017-07-09 um 13:17 schrieb Henri Verbeet:
> Well, kind of. These are actually from the Presentation Forms-B block.
> I'm not sure why those would be especially relevant for the function
> in question though.

I guess Mikhail picked the first one because it looks like a smiley :-).
All chars produce the default character ('?') on my German Windows
installations, which is a reasonable thing to test.

I guess adding some control characters and undefined utf16 values
similarly to the _To_wide tests may be a good idea.